PREDATOR-PREY INTERACTIONS BETWEEN INTRODUCED TROUT AND LONG-TOED SALAMANDERS AND WAYS TO MITIGATE NONCONSUMPTIVE EFFECTS. Erin Kennedy Kenison

Transcription

1 PREDATOR-PREY INTERACTIONS BETWEEN INTRODUCED TROUT AND LONG-TOED SALAMANDERS AND WAYS TO MITIGATE NONCONSUMPTIVE EFFECTS by Erin Kennedy Kenison A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Biological Sciences MONTANA STATE UNIVERSITY Bozeman, Montana May 2014

2 COPYRIGHT by Erin Kennedy Kenison 2014 All Rights Reserved

3 ii ACKNOWLEDGEMENTS I am truly thankful to Dr. Andrea Litt for being an incredible advisor and mentor. She has inspired me to continually strive for success, deepened my enthusiasm for scientific research, and has provided continuous help, support, and encouragement throughout my graduate career. I am truly honored to have been one of her students. I extend my sincerest thanks to my committee members, Drs. David Pilliod and Tom McMahon. It has been wonderful having the opportunity to work with them; their advice, help with study design and research questions, and expert knowledge of amphibians and fish has been exceptional throughout this process. I thank my field technician Michael Forzley, who was invaluable in data collection, excellent company, and continually improved morale in the field during both summer seasons. I thank my fellow graduate students, especially Stacy Davis, Mike Simanonok and Litt Lab members Dan Bachen, Ross Hinderer, and Adam Mitchell for their advice and patient mentoring, rousing scientific discussions, and for always providing necessary laughs. I am grateful to Brian Tornabene for his continuous encouragement, statistical prowess, and the endless supply of happiness he has provided me during this process. I am extremely thankful for my family and their unwavering enthusiasm, patience, and comfort throughout my scientific endeavors. I owe all of my successes to my mother and father - this thesis is dedicated to them. This study was completed with generous financial support from Montana State University, Montana Fish, Wildlife & Parks, Counter Assault, and Montana Institute on Ecosystems.

4 iii TABLE OF CONTENTS 1. INTRODUCTION TO THESIS...1 Objectives...5 Literature Cited NONCONSUMPTIVE EFFECTS OF INTRODUCED TROUT PREDATORS ON LONG-TOED SALAMANDERS: CHANGES IN MORPHOLOGY AND LIFE HISTORY...13 Contribution of Authors and Co-Authors...13 Manuscript Information Page...14 Abstract...15 Introduction...16 Methods...20 Study Sites...20 Sampling...20 Statistical Analyses...22 Results...23 Non-metamorphosing Individuals...23 Metamorphosing Individuals...24 Discussion...24 Literature Cited ADDING VEGETATION STRUCTURE TO REDUCE NONCONSUMPTIVE EFFECTS OF INTRODUCED TROUT: A NOVEL METHOD FOR AMPHIBIAN CONSERVATION?...50 Contribution of Authors and Co-Authors...50 Manuscript Information Page...51 Abstract...52 Introduction...53 Methods...58 Study Sites...58 Field Experiment...58 Statistical Analyses...60 Results...62 Discussion...63 Literature Cited...75

5 iv TABLE OF CONTENTS CONTINUED 4. CONCLUSIONS...82 Literature Cited...85 REFERENCES CITED...87

6 v LIST OF TABLES Table Page 2.1. Locations and characteristics of study lakes with (n = 7) and without trout (n = 7), northwestern Montana, summers 2012 and Factors affecting changes in body morphology of larval salamanders in no trout (n = 177 salamanders in seven lakes) and trout lakes (n = 1372 salamanders in seven lakes), northwestern Montana, summers 2012 and Factors affecting changes in body morphology of larval salamanders over time in no trout (n = 177 salamanders in seven lakes) and trout lakes (n = 1372 salamanders in seven lakes), northwestern Montana, summers 2012 and Allometric relationships, after accounting for time, between weight and SVL, tail length, and tail depth of larval salamanders in no trout (n = 177 salamanders in seven lakes) and trout lakes (n = 1372 salamanders in seven lakes), after accounting for time, northwestern Montana, summers 2012 and Trout effect on body morphology of metamorphosing larval salamanders (means and 95% CIs) in no trout (n = 15 salamanders in seven lakes) and trout lakes (n = 52 salamanders in seven lakes), after accounting for year, northwestern Montana, summers 2012 and Locations and characteristics of study lakes with (n = 2) and without trout (n = 3), northwestern Montana, summer Factors affecting measurements of body morphology of salamanders, 20 per enclosure with repeated measures, in experimental enclosures without trout cues or added structure (n =6 enclosures), without trout cues, but with added structure (n = 6 enclosures), with trout cues, but without added structure (n = 8 enclosures), and with trout cues and added structure (n = 8 enclosures), and northwestern Montana, summer

7 vi LIST OF TABLES CONTINUED Table Page 3.3. Estimated change in salamander body morphology over time (means and 95% CIs) in enclosures with/without trout cues (+/ ) and with/without added structure (+/ ), northwestern Montana, summer Factors affecting body morphology of salamanders at metamorphosis in experimental enclosures without trout cues or structure (n = 18 salamanders in six enclosures), without trout cues, but with added structure (n = 24 salamanders in six enclosures), with trout cues, but without structure (n = 36 salamanders in eight enclosures), and with trout cues and added structure (n = 47 salamanders in eight enclosures), northwestern Montana, summer Factors affecting the probability of metamorphosis in experimental enclosures without trout cues or structure (n = 18 salamanders in six enclosures), without trout cues, but with added structure (n = 24 salamanders in six enclosures), with trout cues, but without structure (n = 36 salamanders in eight enclosures), and with trout cues and added structure (n = 47 salamanders in eight enclosures), northwestern Montana, summer

8 vii LIST OF FIGURES Figure Page 2.1. Map of Montana, where the black rectangle indicates the location of the 14 study lakes, summers 2012 and Measurements of body morphology for larval salamanders: A) SVL, B) total length, C) tail length, and D) tail depth Salamanders captured from a lake without trout (left) and a lake with trout (right), each matches the mean estimates of morphology for each lake type and are photographed at equal scales Body morphology of non-metamorphosing salamanders (means and 95% CIs) in no trout (n = 177) and trout lakes (n = 1372), northwestern Montana, summer 2012 and Changes in body morphology over time of non-metamorphosing salamanders in no trout (n = 177) and trout lakes (n = 1372), northwestern Montana, summer 2012 and Allometric relationships between weight and SVL, tail length, and tail depth for non-metamorphosing salamanders captured in no trout (n = 177) and trout lakes (n = 1372), northwestern Montana, summer 2012 and The probability of metamorphosing salamanders in trout (n = 52 of 1372 salamanders in seven lakes) and no trout lakes (n = 15 of 177 salamanders in seven lakes). Estimates are based on the number of metamorphosing individuals relative to the number of individuals captured over time, northwestern Montana, summer 2012 and Comparisons of body morphology between metamorphosing salamanders (means and 95% CIs) captured in trout (n = 15 salamanders in seven lakes) and no trout lakes (n = 52 salamanders in seven lakes), northwestern Montana, summer 2012 and Vegetation structure constructed for enclosures, made with dead and living plant materials and secured with twine...72

9 viii LIST OF FIGURES CONTINUED Figure Page 3.2. Changes in body morphology of salamanders in enclosures over time in lakes without trout or added structure ( Trout Structure, n = 120 salamanders in six enclosures), without trout, but with added structure ( Trout + Structure, n = 123 salamanders in six enclosures), with trout, but without added structure (+ Trout Structure, n = 160 salamanders in eight enclosures), and with trout and added vegetation structure (+ Trout + Structure, n = 160 salamanders in eight enclosures), northwestern Montana, summer Probability of metamorphosis over time of salamanders in lakes without trout or added structure ( Trout Structure, n = 18 of 120 salamanders), lakes without trout and with added structure ( Trout +Structure, n = 24 of 123 salamanders), lakes with trout, but without added structure (+Trout Structure, n = 36 of 160 salamanders), and lakes with trout and added structure (+Trout +Structure, n = 47 of 160 salamanders), northwestern Montana, summer

10 ix ABSTRACT Predators can increase prey through mortality, but also have the capacity to alter behavior, morphology, and life history through nonconsumptive effects. In many historically fishless lakes in western North America, trout have been introduced for recreational fishing and are associated with reducing and extirpating populations of amphibians, including long-toed salamanders (Ambystoma macrodactylum). Salamanders and trout may coexist in some lakes, as larvae are able to alter foraging behavior by avoiding open water, foraging at night in shallow water, and hiding in cover to avoid predation. However, salamanders may experience nonconsumptive effects due to these behavioral changes. We sought to estimate the nonconsumptive effects of trout on morphology and life history of larval salamanders. We caught salamander larvae using minnow traps in northwestern Montana during the summers of 2012 and 2013 and compared body morphology measurements and size at and timing of metamorphosis between lakes with and without trout. Salamanders in lakes with trout were smaller: they weighed less, had shorter body lengths, and had shorter and narrower tails. Salamanders in lakes with trout were also less likely to metamorphose, did so later in the summer, and had smaller total and tail lengths at metamorphosis. These changes in morphology and life history likely were a result of reduced foraging to avoid predator attacks. We conducted a field experiment in 2013 to investigate whether adding vegetation structure could reduce nonconsumptive effects of trout on salamander larvae by providing refugia and reducing perceived risk of predation. We constructed field enclosures in lakes with and without trout and quantified changes in salamander growth and differences in size at metamorphosis with and without added structure. Salamanders appeared to detect trout cues because they grew more slowly in lakes with trout, even though trout had no ability to consume salamanders. Added vegetation structure did not influence growth rates, but did increase the probability of salamanders that reached metamorphosis. Future research efforts should investigate whether adding vegetation structure to whole lakes can mitigate the nonconsumptive effects of trout, provide a feasible alternative to fish removal, and facilitate coexistence between salamanders and trout.

11 1 CHAPTER ONE INTRODUCTION TO THESIS Currently, amphibians are experiencing drastic declines globally (Stuart et al., 2004). At least 37% of amphibian species are vulnerable, threatened, or endangered and amphibian populations in the United States are estimated to be declining by 3.7% annually (Alford, 2011; Adams et al., 2013). Declines have been attributed to habitat alteration and loss, pollution, contaminants, climate change, increased UV-B radiation, disease, and introductions of nonnative species (Blaustein et al., 1994; Beebee, 1997; Adams, 1999; Collins and Storfer, 2003; Taylor et al., 2005; Pounds et al., 2006). Introduced predators, especially nonnative fish, prey on all life stages of amphibians and have been a leading cause of declines in some areas (Hayes and Jennings, 1986; Hecnar and M Closkey, 1997; Knapp and Matthews, 2000; Pilliod and Peterson, 2001; Denoel et al., 2005; Herwig et al., 2013). In the western United States, fish stocking began in the mid-1900s to increase recreational opportunities for anglers (Bahls, 1992). Introduced trout now exist in 60% of lakes and 95% of all large lakes (>2 ha) historically devoid of fish (Bahls, 1992). Amphibians that breed in historically fishless lakes are more vulnerable to predatory fish, as they often lack chemical or morphological defenses and may be naïve to introduced predators (Kats et al., 1988; Pearl et al., 2003; Welsh et al., 2006; Wells, 2007). Introduced trout have had negative effects on populations of frogs (Lithobates cascadae, Joseph et al., 2011; Rana luteiventris, Pilliod and Peterson, 2001; Rana temporaria,

12 2 Tiberti and Von Hardenberg, 2012; Hyla versicolor, Smith et al., 1999; and Rana muscosa, Wake and Vredenburg, 2008; Pseudacris regilla, Knapp, 2005), newts (Notophthalmus viridescens, Smith et al., 1999; Triturus vulgaris, Aronsson and Stenson, 1995; Taricha torosa, Gamradt and Kats, 1996), and salamanders (Ambystoma gracile, Larson and Hoffman, 2002; Ambystoma macrodactylum, Funk and Dunlap, 1999; Pilliod and Peterson, 2001; Hirner and Cox, 2007; Pearson and Goater, 2008). In the western United States, California yellow-legged frogs (Rana muscosa and Rana sierrae) have been extirpated from >90% of their historic range within the Sierra Nevada Mountains and considerable evidence suggests that the combination of fish stocking and disease have caused these extinctions (Knapp and Matthews, 2000; Vredenburg et al., 2007). Larval populations of the northwestern salamander (Ambystoma gracile) are 12 times smaller in lakes with trout compared to populations in fishless lakes (Larson and Hoffman, 2002). Similarly, Columbia spotted frogs (Rana luteiventris) and long-toed salamanders (Ambystoma macrodactylum) in the Northern Rockies are significantly less abundant in lakes with trout compared to lakes without trout (Pilliod and Peterson, 2001). Long-toed salamanders (Ambystoma macrodactylum) are the most widely distributed amphibian west of the Continental Divide, occupying areas from central California through the Pacific Northwest and extending into southern Alaska (Werner et al., 2004). They are considered to be widespread and abundant by global rankings and are the most common salamander in the state of Montana (Montana State Government, 2014). However, extirpations have occurred in lakes with introduced trout and populations are suspected to be declining despite their secure status within Montana

13 3 (Funk and Dunlap, 1999; Pearson, 2004; Werner et al., 2004; Pilliod et al., 2010; Montana State Government, 2014). Larvae often metamorphose in late summer of the same year they hatch, but may remain in lakes at higher elevations to overwinter or when they require more time to initiate metamorphosis (Wilbur and Collins, 1973). Salamanders are palatable to fish during all aquatic stages and may remain in the aquatic environment for as many as 2-3 years before metamorphosis, making them especially vulnerable to predation (Howard and Wallace, 1985; Welsh et al., 2006). In lakes with trout, survival of larval long-toed salamanders is greatly reduced, populations are less abundant, and the presence of salamanders is negatively correlated with the presence of fish predators (Tyler et al., 1998a; Funk and Dunlap, 1999; Pilliod and Peterson, 2001; Bull and Marx, 2002; Dunham et al., 2004; Welsh et al., 2006; Hirner and Cox, 2007). Salamanders are 44 times less likely to be found in lakes with trout, populations in lakes with trout are 65% less abundant, and the proportion of sites occupied by both trout and salamanders is as low as (Tyler et al., 1998a; Pilliod and Peterson, 2001; Welsh et al., 2006; Hirner and Cox, 2007; Pearson and Goater, 2008). Larval salamanders are capable of detecting fish through chemical and visual cues and alter behavior to reduce encounter rates with fish, most often by increasing refuge use (Tyler et al., 1998b; for other Ambystoma species see: Semlitsch, 1987; Figiel and Semlitsch, 1990; Sih and Kats, 1991, 1994; Storfer and Sih, 1998; Storfer, 1999). Altered behavior can increase salamanders probability of survival, but these changes may come at a cost. Individuals that reduce the amount of time spent foraging in open water may

14 4 decrease energy uptake and body size (Semlitsch, 1987; Figiel and Semlitsch, 1990; Tyler et al., 1998b). Smaller body sizes during the larval period and at metamorphosis are associated with reduced survival and reproduction and increased susceptibility to predation, desiccation, and disease as adults (Petranka et al., 1987; Semlitsch, 1987; Figiel and Semlitsch, 1990; Kiesecker and Blaustein, 1998; Tyler et al., 1998b; Gervasi and Foufopoulos, 2008; Warne et al., 2011). Long-toed salamander populations may be at risk of decline and potential extirpation if trout reduce abundance through predation (consumptive effects) or affect survival and reproductive success of adult salamanders to a point that influences recruitment (nonconsumptive effects) (Kats and Ferrer, 2003; Pilliod et al., 2010). Trout removal has been an effective method to reduce the negative effects of fish predators (Hoffman et al., 2004; Vredenburg, 2004; Knapp et al., 2007; Gross, 2009). Amphibians are able to recolonize sites quickly and increase population size and species diversity (Vredenburg, 2004; Knapp et al., 2007; Gross 2009). Long-toed salamanders were present in five of six lakes where trout had gone extinct, compared to only two of 12 where trout still existed (Funk and Dunlap, 1999). However, this strategy is often unpopular with the public and, like other management efforts (e.g., maintaining buffer zones surrounding breeding sites, creating connectivity between wetlands, or constructing water bodies), can be costly in time, energy, and resources (Semlitsch, 2002; Shulse et al., 2012). Despite the rarity of coexistence, long-toed salamanders and trout have been found to co-occur in lakes where habitat characteristics such as emergent vegetation and

15 5 physical barriers from trout are present (Tyler et al., 1998a; Pearson and Goater, 2008; Pilliod et al., 2010; Pilliod et al., 2013). Structural complexity and available refugia are crucial for larval amphibians to persist with predatory fish, as dense vegetation provides protection, while also reducing the maneuverability, visual range, and effectiveness of fish predators (Werner et al., 1983; McNair, 1986; Sih, 1987; Kats et al., 1988; Sih et al., 1988; Figiel and Semlitsch, 1991; Diehl, 1992; Babbitt and Jordan, 1996; Babbitt and Tanner, 1997; Tyler et al., 1998b; Allouche, 2002; Pearson, 2004). However, we are not aware of studies that have experimentally manipulated cover or added structural complexity as a method to enhance amphibian coexistence with trout predators. Understanding the nonconsumptive effects of introduced trout on long-toed salamanders, the ways in which predator-prey dynamics affect morphology and life history of prey species, and how these effects contribute to prey persistence and coexistence will be important in designing new and effective conservation strategies that help preserve salamander populations without requiring the removal of trout. Objectives: We aimed to examine the nonconsumptive effects introduced trout have on longtoed salamander larvae and test the effectiveness of a new conservation strategy intended to facilitate coexistence between these species. Chapter two is focused on understanding predator-prey interactions and the nonconsumptive effects introduced trout have on long-toed salamander larvae. Specifically, we will investigate morphology, before and at metamorphosis, of

16 6 salamanders in lakes with and without trout. We will also examine the influence of trout on life history, namely the probability and timing of metamorphosis. Chapter three investigates the efficacy of adding complex vegetation structure to reduce the nonconsumptive effects of trout on larval salamanders. We seek to test the effectiveness of this management strategy and its potential use in conservation efforts for other amphibian populations negatively affected by introduced predators.

17 7 Literature Cited Adams, M Correlated factors in amphibian decline: exotic species and habitat change in western Washington. The Journal of Wildlife Management 63: Adams, M.J., D.A.W. Miller, E. Muths, P.S. Corn, E.H. Campbell Grant, L.L. Bailey, G.M. Fellers, R.N. Fisher, W.J. Sadinski, H. Waddle, and S.C. Walls Trends in amphibian occupancy in the United States. PLOS ONE 8:1-5. Alford, R.A Bleak future for amphibians. Nature 480: Allouche, S Nature and functions of cover for riverine fish. Bulletin Français de la Pêche et de la Pisciculture 365/366: Aronsson, S. and J.A.E. Stenson Newt-fish interactions in a small forest lake. Amphibia-Reptilia 16: Babbitt, K.J. and F. Jordon Predation on Bufo terrestris tadpoles, effects of cover and predator identity. Copeia 2: Babbitt, K.J. and G.W. Tanner Effects of cover and predator identity on predation of Hyla squirella tadpoles. Journal of Herpetology 31: Bahls, P.F The status of fish populations and management of high mountain lakes in the western United States. Northwest Science 66: Beebee, T.J.C Changes in dewpond numbers and amphibian diversity over 20 years on chalk downland in Sussex, England. Biological Conservation 81: Blaustein, A.R., P.D. Hoffman, D.G. Hokit, J.M. Kiesecker, S.C. Walls, and J.B. Hays UV repair and resistance to solar UV-B in amphibian eggs: a link to population declines? Proceedings of the National Academy of Sciences 91: Bull, E.L. and D.B. Marx Influence of fish and habitat on amphibian communities in high elevation lakes in northeastern Oregon. Northwest Science 76: Collins, J.P. and A. Storfer Global amphibian declines: sorting the hypotheses. Diversity and Distribution 9:89-98.

18 8 Denoel, M., G. Dzukic, and M.L. Kalezic Effects of widespread fish introductions on paedomorphic newts in Europe. Conservation Biology 19: Diehl, S Fish predation and benthic community structure: the role of omnivory and habitat complexity. Ecology 73: Dunham, J.B., D.S. Pilliod, and M.K. Young Assessing the consequences of nonnative trout in headwater ecosystems in western North America. Fisheries 29: Figiel, C.R., Jr. and R.D. Semlitsch Population variation in survival and metamorphosis of larval salamanders (Ambystoma maculatum) in the presence and absence of fish predation. Copeia 1990: Figiel, C.R. Jr. and R.D. Semlitsch Effects of nonlethal injury and habitat complexity on predation in tadpole populations. Canadian Journal of Zoology 69: Funk, W.C. and W.W. Dunlap Colonization of high-elevation lakes by long-toed salamanders (Ambystoma macrodactylum) after the extinction of introduced trout populations. Canadian Journal of Zoology 77: Gamradt, S.C. and L.B. Kats Effects of introduced crayfish and mosquitofish on California newts. Conservation Biology 10: Gervasi, S.S. and J. Foufopoulos Costs of plasticity: responses to desiccation decrease post-metamorphic immune function in a pond-breeding amphibian. Functional Ecology 22: Gross, L.M Long-term assessment of predatory fish removal on a pond-breeding amphibian community in central Illinois. Masters Thesis. Eastern Illinois University, Charleston, IL. Hayes, M.P. and M.R. Jennings Decline of ranid frog species in western North America: are bullfrogs (Rana catesbeiana) responsible? Journal of Herpetology 20: Hecnar, S.J. and R.T. M Closkey The effects of predatory fish on amphibian species richness and distribution. Biological Conservation 79: Herwig, B.R., L.W. Schroeder, K.D. Zimmer, M.A. Hanson, D.F. Staples, R.G. Wright, and J.A. Younk Fish influences on amphibian presence and abundance in prairie and parkland landscapes of Minnesota, USA. Journal of Herpetology 47:

19 9 Hirner, J.L.M. and S.P. Cox Effects of rainbow trout (Oncorhynchus mykiss) on amphibians in productive recreational fishing lakes of British Columbia. Canadian Journal of Fish and Aquatic Sciences 64: Howard, J.H. and R.L. Wallace Life history characteristics of populations of longtoed salamander (Ambystoma macrodactylum) from different altitudes. American Midland Naturalist 113: Joseph, M.B., J. Piovia-Scott, S.P. Lawler, and K.L. Pope Indirect effects of introduced trout on Cascades frogs (Rana cascadae) via shared aquatic prey. Freshwater Biology 56: Kats, L.B., J.W. Petranka, and A. Sih Antipredator defenses and the persistence of amphibian larvae with fishes. Ecology 69: Kats, L.B. and R.P. Ferrer Alien predators and amphibian declines: review of two decades of science and the transition to conservation. Diversity and Distributions 9: Kiesecker, J.M. and A.R. Blaustein Effects of introduced bullfrogs and smallmouth bass on microhabitat use, growth, and survival of native red-legged frogs (Rana aurora). Conservation Biology 12: Knapp, R.A. and K.R. Matthews Non-native fish introductions and the decline of the mountain yellow-legged frog from within protected areas. Conservation Biology 14: Knapp, R.A Effects of nonnative fish and habitat characteristics on lentic herpetofauna in Yosemite National Park, USA. Biological Conservation 121: Knapp, R.A., D.M. Boiano, and V.T. Vredenburg Removal of nonnative fish results in population expansion of a declining amphibian (mountain yellowlegged frog, Rana muscosa). Biological Conservation 135: Larson, G.L. and R.L. Hoffman Abundances of northwestern salamander larvae in montane lakes with and without fish, Mount Rainer National Park, Washington. Northwest Science 76: McNair, J.N The effects of refuges on predator-prey interactions: a reconsideration. Theoretical Population Biology 29: Montana State Government, Montana Field Guide. Available at:

20 10 Pearl, C.A., M.J. Adams, G.S. Schuytema, and A.V. Nebeker Behavioral responses of anuran larvae to chemical cues of native and introduced predators in the Pacific northwestern United States. Journal of Herpetology 37: Pearson, K.J The effects of introduced fish on the long-toed salamander (Ambystoma macrodactylum) in southwestern Alberta, Canada. Masters Thesis. University of Lethbridge, Lethbridge, Alberta, Canada. Pearson, K.J. and C.P. Goater Distribution of long-toed salamanders and introduced trout in high- and low-elevation wetlands in southwestern Alberta, Canada. Ecoscience 15: Petranka, J.W., L.B. Kats, and A. Sih Predator-prey interactions among fish and larval amphibians: use of chemical cues to detect predatory fish. Animal Behaviour 35: Pilliod, D.S. and C.R. Peterson Local and landscape effects of introduced trout on amphibians in historically fishless watersheds. Ecosystems 4: Pilliod, D.S., B.R. Hossack, P.F. Bahls, E.L. Bull, and P.S. Corn Non-native salmonids affect amphibian occupancy at multiple spatial scales. Diversity and Distributions 16: Pilliod, D.S., R.S. Arkle, and B.A. Maxell Persistence and extirpation in invaded landscapes: patch characteristics and connectivity determine effects of non-native predatory fish on native salamanders. Biological Invasions 15: Pounds, J.A., M.R. Bustamante, L.A. Coloma, J.A. Consuegra, M.P.L. Fogden, P.N. Foster, E. La Marca, K.L. Masters, A. Merino-Viteri, R. Puschendorf, S.R. Ron, G.A. Sanchez-Azofeifa, C.J. Still, and B.E. Young Widespread amphibian extinctions from epidemic disease driven by global warming. Nature 439: Semlitsch, R.D Interactions between fish and salamander larvae. Costs of predator avoidance or competition? Oecologia 72: Semlitsch, R.D Critical elements for biologically based recovery plans of aquaticbreeding amphibians. Conservation Biology 16: Shulse, C.D., R.D. Semlitsch, K.M. Trauth, and J.E. Gardner Testing wetland features to increase amphibian reproductive success and species richness for mitigation and restoration. Ecological Applications 22:

21 Sih, A Prey refuges and predator-prey stability. Theoretical Population Biology 31: Sih, A., J.W. Petranka, and L.B. Kats The dynamics of prey refuge use: a model and tests with sunfish and salamander larvae. The American Naturalist 132: Sih, A. and L.B. Kats Effects of refuge availability on the responses of salamander larvae to chemical cues from predatory green sunfish. Animal Behaviour 42: Sih, A. and L.B. Kats Age, experience, and the response of streamside salamander hatchlings to chemical cues from predatory sunfish. Ethology 96: Smith, G.R., J.E. Rettig, G.G. Mittelbach, J.L. Valiulis, and S.R. Schaack The effects of fish on assemblages of amphibians in ponds: a field experiment. Freshwater Biology 41: Storfer, A. and A. Sih Gene flow and ineffective antipredator behavior in a streambreeding salamander. Evolution 52: Storfer, A Gene flow and local adaptation in a sunfish-salamander system. Behavioral Ecology and Sociobiology 46: Stuart, S.N., J.S. Chanson, N.A. Cox, B.E. Young, A.S.L. Rodrigues, D.L. Fischman, and R.W. Waller Status and trends of amphibian declines and extinctions worldwide. Science 306: Taylor, B., D. Skelly, L.K. Demarchis, M.D. Slade, D. Galusha, and P.M. Rabinowitz Proximity to pollution sources and risk of amphibian limb malformation. Environmental Health Perspectives 113: Tiberti, R and A. Von Hardenberg Impact of introduced fish on common frog (Rana temporaria) close to its altitudinal limit in alpine lakes. Amphibia-Reptilia 33: Tyler, T., W.J. Liss, L.M. Ganio, G.L. Larson, R. Hoffman, E. Deimling, and G. Lomnicky. 1998a. Interaction between introduced trout and larval salamanders (Ambystoma macrodactylum) in high-elevation lakes. Conservation Biology 12: Tyler, T.J., W.J. Liss, R.L. Hoffman, and L.M. Ganio. 1998b. Experimental analysis of trout effects on survival, growth, and habitat use of two species of Ambystomatid salamanders. Journal of Herpetology 32:

22 12 Vredenburg, V.T Reversing introduced species effects: experimental removal of introduced fish leads to rapid recovery of declining frog. Proceedings of the National Academy of Sciences 101: Vredenburg, V.T., R. Bingham, R. Knapp, J.A.T. Morgan, C. Moritz, and D. Wake Concordant molecular and phenotypic data delineate new taxonomy and conservation priorities for the endangered mountain yellow-legged frog. Journal of Zoology 271: Wake, D.B. and V.T. Vredenburg Are we in the midst of the sixth mass extinction? A view from the world of amphibians. Proceedings of the National Academy of Sciences 105: Warne, R.W., E.J. Crespi, and J.L. Brunner Escape from the pond: stress and developmental responses to ranavirus infection in wood frog tadpoles. Functional Ecology 25: Wells, K.D The Ecology & Behavior of Amphibians. The University of Chicago Press, United States. Welsh, H.H., Jr., K.L. Pope, and D. Boiano Sub-alpine amphibian distributions related to species palatability to non-native salmonids in the Klamath mountains of northern California. Diversity and Distributions 12: Werner, E.E., J.F. Gilliam, D.J. Hall, and G.G. Mittelbach An experimental test of the effects of predation risk on habitat use in fish. Ecology 64: Werner, J.K., B.A. Maxell, P. Hendricks, and D.L. Flath Amphibians and reptiles of Montana. Mountain Press Publishing, United States. Wilbur, H.M. and J.P. Collins Ecological aspects of amphibian metamorphosis. Science 182:

23 13 CHAPTER TWO NONCONSUMPTIVE EFFECTS OF INTRODUCED TROUT PREDATORS ON LONG-TOED SALAMANDERS: CHANGES IN MORPHOLOGY AND LIFE HISTORY Contributions of Authors and Co-Authors Manuscript in Chapter 2 Author: Erin K. Kenison Contributions: Conceived the study, collected and analyzed data, and authored the manuscript. Co-author: Andrea R. Litt Contributions: Conceived the study, assisted with study design, and obtained funding. Discussed results, aided in interpretation of analyses, and edited manuscript. Co-author: David S. Pilliod Contributions: Helped conceive the study and assisted with study design. Discussed interpretation and implications of results and reviewed manuscript. Co-author: Thomas E. McMahon Contributions: Helped conceive the study and assisted with study design. Discussed interpretation and implications of results and reviewed manuscript.

24 14 Manuscript Information Page Erin K. Kenison, Andrea R. Litt, David S. Pilliod, and Thomas E. McMahon Journal of Herpetology Status of Manuscript: x_ Prepared for submission to a peer-reviewed journal Officially submitted to a peer-reviewed journal Accepted by a peer-reviewed journal Published in a peer-reviewed journal Published by The Society for the Study of Amphibians and Reptiles

25 15 CHAPTER TWO NONCONSUMPTIVE EFFECTS OF INTRODUCED TROUT PREDATORS ON LONG-TOED SALAMANDERS: CHANGES IN MORPHOLOGY AND LIFE HISTORY Abstract Predators can influence prey directly through consumption or by altering prey behavior, morphology, and life history. Facultative changes to reduce predation often come with a cost, such as expenditure of energy and resources, reduced size, and potentially reduced fecundity and survival. In western North America, trout have been introduced to lakes for recreational fishing, leading to reductions and extirpations of long-toed salamanders (Ambystoma macrodactylum). Salamanders and trout may coexist in some lakes, as larvae are able to alter foraging activity and time in refugia. However, salamanders may experience nonconsumptive effects due to these changes in behavior. We sought to investigate the effects of trout on salamander morphology and life history We sampled lakes with and without trout in northwestern Montana during the summers of 2012 and We captured salamander larvae using minnow traps and compared morphological measurements of non-metamorphosing and metamorphosing salamanders between lakes with and without trout. Salamanders in lakes with trout were smaller: they weighed less, had shorter body lengths, and had shorter and narrower tails. Salamanders in lakes with trout were >20 times less likely to metamorphose and those that did, metamorphosed later and had smaller total and tail lengths. Changes in body morphology

26 16 and life history were likely a result of predator-avoidance behavior, such as increased vigilance, reduced foraging, and increased refuge use. Defensive strategies may reduce predation, but may affect reproduction and survival in the long term. Future research should be directed at understanding how nonconsumptive effects experienced during the larval stage translate to survival and reproduction of adult salamanders. Introduction Predator-prey interactions are viewed most often in terms of consumption (Mittelbach, 2012). However, predator-prey interactions are much more diverse and complex when nonlethal, nonconsumptive effects also are considered (Huang and Sih, 1991; Beckerman et al., 1997; Werner and Peacor, 2003; Joseph et al., 2011). To avoid predation, prey use tactile, visual, or chemical cues to detect predators and respond with appropriate defensive tactics to increase their probability of survival (Kats et al., 1988; Stauffer and Semlitsch, 1993; Lima, 1998). When defenses are not innate (e.g., unpalatable toxins in skin of newts and toads, Daly, 1995), the presence of predators may induce facultative defenses that result in marked changes in prey behavior, morphology, and life history (Gilliam and Fraser, 1987; Ludwig and Rowe, 1990; Brönmark and Miner, 1992; McNamara and Houston, 1994; Lima, 1998; Peckarsky et al., 2001, 2002; Benard, 2004; McPeek, 2004; Mittelbach, 2012). Although such defensive strategies may reduce the risk of predation, they also can reduce prey energy intake, reproductive success, and survival (Gilliam, 1982; Werner et al., 1983; Gilliam and Fraser, 1987; Ludwig and Rowe, 1990; McNamara and Houston,

27 ; Lima, 1998; McPeek and Peckarsky, 1998; Peckarsky et al., 2002; Benard, 2004). For example, when small, bluegill sunfish (Lepomis macrochirus) are exposed to predators, they increase foraging activity in vegetated areas that are resource-poor relative to open areas that are resource-rich, which results in a 27% reduction in growth (Werner et al., 1983). Elk (Cervus elaphus) spend more time in protected, coniferous forest and less time in grasslands that are rich in foraging resources when wolves are present, which reduces fecundity by 35% (Creel et al., 2005). Similarly, larval dragonflies reared with caged predators have higher levels of mortality and metamorphic failure, despite predators inability to actually capture and consume prey (McCauley et al., 2011). Thus, predator avoidance tactics confer benefits by reducing predation risk, but prey my may incur costs to growth and survival with such responses (Brönmark and Hansson, 2005). The standard interplay between predators and prey becomes more complex when predators become established in novel environments (Salo et al., 2007). Unlike prey that have coexisted with native predators over evolutionary time, naïve prey may lack chemical or morphological defenses, be unable to detect predators as a threat, or both (Kats et al., 1988; Kiesecker and Blaustein, 1997; Preisser et al., 2005; Salo et al., 2007). Even prey that are capable of assessing predation risk and responding with altered behavior have reduced growth, are smaller in size, and have higher rates of mortality compared to prey in the absence of introduced predators (Tyler et al., 1998a; Pearson and Goater, 2009). Nonconsumptive effects can be as great or greater than consumptive effects, and together, can influence prey abundance and predator-prey dynamics (Macan,

28 ; Werner and Peacor, 2003; Bolnick and Preisser, 2005; Preisser et al., 2005; Finlay and Vredenburg, 2007). In the western United States, trout have been introduced for recreational fishing to 95% of historically fishless lakes and have replaced native amphibians as top predators within some of these lakes (Bahls, 1992; Dunham et al., 2004). Long-toed salamanders (Ambystoma macrodactylum) are considered the most common salamander in Montana, yet populations have decreased and been extirpated in some lakes with introduced trout (Funk and Dunlap, 1999; Pilliod et al., 2010). Salamanders are palatable to trout and are susceptible to predation during all developmental stages from egg to adult (Tyler et al., 1998a; Pearson and Goater, 2009). These salamanders require deep, permanent lakes for overwintering, oviposition, and development during their 1-3 year larval period and infrequently coexist with trout (Tyler et al., 1998b; Funk and Dunlap, 1999; Pilliod and Peterson, 2001; Bull and Marx, 2002; Pearson and Goater, 2008; Pilliod et al., 2010, 2013). In lakes with trout, survival of long-toed salamander larvae is greatly reduced, populations are less abundant, and the presence of salamanders is negatively correlated with the presence of fish predators (Tyler et al., 1998b; Funk and Dunlap, 1999; Pilliod and Peterson, 2001; Bull and Marx, 2002; Dunham et al., 2004; Welsh et al., 2006; Hirner and Cox, 2007). The rarity of coexistence and reduced abundance of salamanders in lakes with trout is primarily thought to be a result of direct predation. Long-toed salamander larvae are capable of detecting chemical and visual cues of fish predators and use behavioral strategies to avoid predation (Tyler et al., 1998a; Pearson and Goater, 2009). Larvae often increase the use of cover objects or occur in

29 19 dense vegetation or shallow water, which reduces encounter rates with trout and decreases the maneuverability and visual range of these predators (Werner et al., 1983; Figiel and Semlitsch, 1991; Babbitt and Jordan, 1996). However, increased refuge use is associated with reduced foraging activity (Petranka et al., 1987; Skelly and Werner, 1990; Pearl et al., 2003; Paoletti et al., 2011). Spending more time in refugia can limit the amount of resources acquired and negatively influence growth and size (Petranka et al., 1987; Skelly and Werner, 1990; Tyler et al., 1998b; Nyström et al., 2001; Benard, 2004; Davenport et al., 2013). Predator-induced changes in behavior also can correspond to changes in life history (Benard, 2004). Increased predation risk in the larval stage may encourage early metamorphosis, so individuals can minimize their exposure to predators and escape environments with high risk of mortality (Ludwig and Rowe, 1990; Rowe and Ludwig, 1991; Abrams and Rowe, 1996; Chivers et al., 1999; Lardner, 2000). Alternatively, antipredator defenses that result in reduced growth rates may extend larval periods and require more time for larvae to acquire necessary resources for metamorphosis (Wilbur and Collins, 1973; Lardner, 2000; Barnett and Richardson, 2002; Benard, 2004; Skelly, 1992). We aimed to examine the nonconsumptive effects of introduced trout on longtoed salamander morphology and timing and size at metamorphosis. We used lakes with and without trout as a model system to better understand predator-prey interactions and the nonconsumptive effects incurred by prey. We hypothesized that salamanders in lakes

30 20 with trout would be smaller in size and that the presence of trout would shorten the larval period of salamanders. Methods Study Sites: Our study took place in permanent lakes in northwestern Montana between the cities of Hungry Horse and Helena, Montana (Figure 2.1). We sampled lakes primarily northeast of Missoula, in the South Fork Flathead River and Swan River drainages; elevation ranged from 953 to 1,988 m and lake perimeters ranged from 140 to 960 m (Table 2.1). In 2012, we sampled 14 lakes for long-toed salamanders, seven lakes with trout and seven without trout. In 2013, we sampled for salamanders in six of the original 14 lakes, three with trout and three without, to increase our trapping effort within each lake. We did not randomly select sites; we considered lakes where previous surveys verified the presence of salamanders, salamanders and trout (B. Maxell, pers. comm.; M. Boyer, pers. comm.), or lakes of similar size that we found with local maps. We identified trout as rainbow trout (Oncorhynchus mykiss), cutthroat trout (Oncorhynchus clarkia), or hybrids of the two species based on field observations, recreating anglers, and Montana stocking reports (Montana Fish, Wildlife & Parks, 2004, 2005, 2007, 2010, 2011). We selected lakes for our study based on visual encounter and trapping surveys in 2012 to confirm the presence of salamanders and the presence or absence of trout. Sampling: We used minnow traps to capture salamanders as they are especially useful to trap

31 21 small, cryptic larvae, are associated with little mortality or injury, and capture individuals passively without bait (Adams et al., 1997). We set all traps in the littoral zone around the perimeter of each lake at depths 1 m. In 2012, the total number of traps in each lake ranged from 6 to 16 and in 2013, the total number of traps in each lake ranged from 7 to 48. In 2012, we collected measurements of dissolved oxygen, ph, and water temperature using a YSI Pro 1030 meter at each study lake, during each visit, close to the lake shore in water <1 m deep. We deployed traps for four-day periods from July through August of 2012 and We visited lakes a total of four times in 2012 and three times in During each visit, we collected salamanders from traps, anesthetized individuals with MS-222, measured weight, snout-vent length (SVL), total length, tail length, and tail depth (Figure 2.2) with an electronic scale and calipers, and marked individuals with visual implant elastomer in the middle of the tail. We marked individuals uniquely except when we captured >100 salamanders during one visit to a lake, we used batch marks specific to the trap and visit number. Elastomer is a common marking technique for larval salamanders, is long lasting, does not affect survival or metamorphosis, and allowed us to identify individual salamanders if and when we recaptured them (Grant, 2008). We also noted whether individuals had initiated metamorphosis, defined as showing any evidence of gill absorption (Dodd and Dodd, 1976; Arntzen, 1981; Duellman and Trueb, 1986). We released all individuals at their original capture location following data collection (Montana State University Institutional Animal Care and Use Committee protocol ).

32 22 Statistical Analyses: We compared body morphology (weight, SVL, total length, tail length, tail depth, and tail index) of larvae between lakes with and without trout and made separate comparisons for larvae that were metamorphosing and those that were not. We calculated tail index by dividing tail length by SVL to estimate tail size relative to the body (Arntzen, 1981). We evaluated allometric relationships between weight and SVL, tail length, and tail depth of non-metamorphosing salamanders and compared slopes of these relationships between lakes with and without trout (Arntzen, 1994; Delgado-Acevedo and Restrepo, 2008). We also compared the probability and timing of salamander metamorphosis in lakes with and without trout; these estimates were based on the number of individuals exhibiting signs of metamorphosis out of the total number of individuals captured at each visit. Lastly, we computed the total number of salamanders captured divided by lake perimeter and made comparisons between lakes with and without trout, to assess differences in salamander density. We used a generalized linear mixed model approach for all analyses, selecting the appropriate distribution and link function for each response variable. We treated lakes as our subjects and included a three-tiered, nested data structure multiple visits within lakes within years as random effects in all analyses to account for repeated measurements and variation among lakes (Zuur et al., 2009). We included the presence or absence of trout (trout) as the main effect in our analytical model. We also included year and number of days since the first sampling visit to account for changes over time and to test for interactive effects (trout*day). We removed interactions that did not explain

33 23 sufficient variation (α >0.1) from models for inference, but kept all simple explanatory variables (i.e., trout, year, and day). We included lake level covariates (i.e., perimeter, elevation, and repeated measures of dissolved oxygen, ph, and temperature) to better account for variation among lakes. We log-transformed response variables when appropriate to meet assumptions. We ran all statistical analyses in program R version and used packages nlme and MASS (R Developmental Core Team, 2008). We report mean values, percent differences, and 95% confidence intervals in text and tables; we back-transformed these values when appropriate. Results Non-metamorphosing Individuals: Salamander densities were similar between lakes with and without trout, after accounting for year (estimated difference = 6 salamanders/100 m, 95% CI = salamanders/100 m, t 12 = -0.45, P = 0.66). Salamanders were consistently smaller in size in lakes with trout compared to lakes without trout and only some measurements changed over the sampling season (Figure 2.3, Tables 2.2 and 2.3). In lakes with trout, salamander larvae weighed 38% less (95% CI = 3 to 72%), were 24% shorter (6 to 43%) in total length, and had tails 29% shorter (7 to 50%, Figure 2.4); these differences were present at the beginning of the summer and consistent throughout the sampling season. Salamanders in lakes with trout were 19% shorter (-1 to 38%) based on SVL and although SVL of salamanders increased as they grew over the sampling season, these differences persisted (Figure 2.5, Table 2.3). Tail depth changed over time and differences depended on the

34 24 presence or absence of trout; salamanders in lakes with trout had tails 24% narrower (7 to 41%) at the beginning of the summer, but had comparable tail depths near the end of the summer (Figure 2.5, Table 2.3). Tail index decreased over time and salamanders in lakes with trout had tail length to SVL ratios 0.06 less (0.03 to 0.09, Figure 2.5, Table 2.3). Measurements of body morphology were similar between years (Table 2.2, 2.3) Allometric relationships between weight and SVL, tail length, and tail depth were positive and linear, and the slopes differed between lakes with and without trout (trout * weight, Figure 2.6, Table 2.4). Metamorphosing Individuals: Larval salamanders were >20 times more likely to metamorphose in lakes without trout and they metamorphosed earlier in the summer, compared to larvae in lakes with trout (t 1553 = -2.25, P = 0.025, Figure 2.7). Individuals that showed signs of metamorphosis in lakes with trout were 8.8% shorter in total length (95% CI = 3 to 14.9%, t 6 = -2.98, P = 0.02) and 16.9% shorter in tail length (7.4 to 26%, t 6 = 3.59, P = 0.01) than salamanders in lakes without trout (Figure 2.8). There was no difference in weight, SVL, and tail depth measurements between metamorphosing salamanders in lakes with and without trout (Table 2.5). Discussion Predation, especially by nonnative predators, can reduce or extirpate entire populations of prey (Heyer et al., 1975). Prey may be able to mitigate consumptive

35 25 effects temporarily, yet they are continually shifting between safe and unsafe states when predators are present (Sih, 1987). By altering the behavior of prey species, through changes in diet or habitat selection, predators also can induce nonconsumptive effects (Huang and Sih, 1991; Schmitz, 1998). Although amphibians occasionally are able to exist with introduced predatory fish, our findings demonstrate that nonconsumptive effects of trout on long-toed salamanders impact larval morphology and timing of metamorphosis. Salamanders that increase their use of vegetated shallows when predators are detected can successfully coexist with trout, but these behavioral modifications come with costs (Anderson and Williamson, 1974; Taylor, 1983; Stenhouse, 1985; Sih et al., 1988; Babbitt and Jordan, 1996; Tyler et al., 1998b; Van Buskirk and Schmidt, 2000). Salamanders existing in lakes with trout were consistently smaller in size (Figure 2.3). With as much as a 38% reduction in weight and a 24% reduction in total length; these decreases are comparable to California red-legged frog tadpoles (Rana aurora draytonii) that weighed 34% less at metamorphosis in the presence of mosquitofish (Gambusia affinis) and larval mole salamanders (Ambystoma talpoideum) that had 18% reductions in body size in the presence of bluegill sunfish (Lepomis macrochirus) in constructed ponds (Semlitsch, 1987; Lawler et al., 1999). Furthermore, variation in allometry relating weight with body and tail measurements suggests different biological scaling relationships for salamanders during larval development. More specifically, in lakes with trout, salamanders increase body and tail shape and size more slowly with increasing weight compared to larvae in lakes without trout. Unless a small size is advantageous to

36 26 larval salamanders and reduces detection by trout, increases potential refugia in compact spaces, or is genetically selected for in lakes with trout, small sizes likely are a consequence of reduced activity and feeding to avoid predator attacks (Tyler et al., 1998a; Yurewicz, 2004). Larvae in environments with trout must evaluate the tradeoff between predatoravoidance and procuring resources, which can influence size and timing of metamorphosis (Wilbur and Collins, 1973; Duellman and Trueb, 1986; Semlitsch, 1987; Skelly, 1995; Tyler et al., 1998b; Hirner and Cox, 2007). Salamanders in lakes with trout were less likely to metamorphose and did so later in the summer season. Amphibian metamorphosis is initiated once a minimum body size is attained (Wilbur and Collins, 1973). The significantly smaller size of non-metamorphosing larvae in lakes with trout may have reduced probability of metamorphosis and restricted salamanders to the aquatic environment for longer periods of time, in order to reach adequate sizes for metamorphosis (Wilbur and Collins, 1973; Arntzen, 1981; Bull and Marx, 2002). Much like long-toed salamanders, the larval period of wood frogs (Lithobates sylvaticus) is significantly longer in the presence of predatory fish (Davenport et al., 2013). Lengthening the larval periods can increase exposure and susceptibility to trout predation, especially during winter months when larvae must move away from shallow, vegetated areas along the lake perimeter and into open, deep sections of the lake that do not freeze (Emery et al., 1972; Wilbur and Collins, 1973; Scott, 1990). We did not recapture any individuals marked in 2012 in the summer of 2013 and others only have observed second-year larvae in lakes without fish, which may provide evidence of increased

37 27 predation of individuals that attempt to overwinter (Bull and Marx, 2002). Timing and size at metamorphosis can also be density dependent and driven by availability of food or permanency of habitat (Wilbur and Collins, 1973; Duellman and Trueb, 1986). Increased density is correlated with smaller size and extended larval periods in permanent bodies of water, whereas reduced resources are related to smaller size and accelerated development (Semlitsch and Caldwell, 1982; Petranka, 1989; Scott, 1990; Van Buskirk and Smith, 1991; Newman, 1994). We did not detect differences in salamander density between lakes with and without trout, but the probability and timing of metamorphosis of salamanders was delayed significantly in lakes with trout. Relatively small differences in size at metamorphosis can affect reproduction, as the number and size of eggs are positively correlated with size of females (Kaplan and Salthe, 1979; Berven, 1988; Davenport et al., 2013). Mayflies emerge earlier from streams to avoid trout predation, yet, they are 13-20% smaller in size and have a 24-35% reduction in fecundity (McPeek and Peckarsky, 1998; Peckarsky et al., 2001, 2002). A decrease of 3 mm in body length of wood frogs results in a decrease of 90 eggs, which represents a 15% decrease in the mean clutch size (Berven, 1982, 1988). Salamanders that metamorphosed in lakes with trout had smaller total and tail sizes compared to salamanders in lakes without trout. Tail length, in addition to body size, is positively related to clutch size of female golden-striped salamanders (Chioglossa lusitanica), as tails serve as food storage and may provide resources for reproduction (Maiorana, 1977; Arntzen, 1981, 1994). Salamanders reproduce as early as one year after metamorphosing from lakes, which provides limited opportunity to compensate for reduced total and tail

38 28 sizes observed at metamorphosis (Berven, 1988; Boone et al., 2007). Moreover, adult salamanders are able to detect the presence of fish predators and are less likely to oviposit eggs in lakes with trout; if they do, abundance of egg masses, and therefore eggs, are significantly lower compared to lakes without trout, further reducing reproductive success in a population (Bull and Marx, 2002). Nonconsumptive effects of trout clearly affect larval size, timing of metamorphosis, and size at metamorphosis. Reduced body sizes and delayed metamorphosis of salamanders in lakes with trout provide clear expressions of the morphological and metamorphic costs associated with coexisting with predators. Although remaining in refugia will reduce encounter rates with predators, the resulting changes in body morphology and life history are correlated with lower survival (Smith, 1987; Semlitsch et al., 1988; Beck and Congdon, 1999; Altwegg and Reyer, 2003; Orizaola and Brana, 2005). In addition, the stress associated with predator-induced behaviors can suppress immune function and make amphibians more vulnerable to disease, pathogens, or parasites (Stuart et al., 2004; Gervasi and Foufopoulos, 2008; Warne et al., 2011). We understand little about whether the nonconsumptive effects of trout we documented persist to affect survival and reproduction of adult salamanders, however small size and increased stress during the larval period are related directly to smaller size at metamorphosis, which can decrease adult survival, mating success, and reproductive potential (Petranka et al., 1987; Semlitsch, 1987; Figiel and Semlitsch, 1990; Kiesecker and Blaustein, 1998; Boone et al., 2007). Further research focusing on the costs prey

39 29 species incur to coexist with predators could provide better understanding of predatorprey interactions, the importance of nonconsumptive effects relative to consumptive effects, and how introduced predators may influence prey populations.

40 30 Tables Table 2.1: Locations and characteristics of study lakes with (n = 7) and without trout (n = 7), northwestern Montana, summers 2012 and 2013 (! denotes lakes sampled in 2012 and 2013) Lake Latitude Longitude Trout Perimeter (m) Elevation (m) Salamanders/ 100m Alpha No Elk! No LL No LL No Loveboat No New! No Pee! No Beta Yes Elsina! Yes Horseshoe Yes Rock! Yes Sink Yes Spook! Yes Telegraph Yes

41 Table 2.2: Factors affecting changes in body morphology of larval salamanders over time in no trout (n = 177 salamanders in seven lakes) and trout lakes (n = 1372 salamanders in seven lakes), northwestern Montana, summers 2012 and 2013 Weight Total Length Tail Length Explanatory df Estimate t P Estimate t P Estimate t P Trout Day Year Table 2.3: Factors affecting changes in body morphology of larval salamanders over time in no trout (n = 177 salamanders in seven lakes) and trout lakes (n = 1372 salamanders in seven lakes), northwestern Montana, summers 2012 and 2013 SVL Tail Depth Tail Index Explanatory df Estimate t P Estimate t P Estimate t P Trout Day <0.001 Year Trout * Day

42 32 Table 2.4: Allometric relationships, after accounting for time, between weight and SVL, tail length, and tail depth of larval salamanders in no trout (n = 177 salamanders in seven lakes) and trout lakes (n = 1372 salamanders in seven lakes), after accounting for time, northwestern Montana, summers 2012 and 2013 SVL Tail Length Tail Depth Explanatory df t P t P t P Trout Weight < < <0.001 Trout* Weight <0.001

43 Table 2.5: Trout effect on body morphology of metamorphosing larval salamanders (means and 95% CIs) in no trout (n = 15 salamanders in seven lakes) and trout lakes (n = 52 salamanders in seven lakes), after accounting for year, northwestern Montana, summers 2012 and 2013 Weight SVL Total Length Tail Length Tail Depth Tail Index Explanatory df t P t P t P t P t P t P Trout Year

44 34 Figures Figure 2.1: Map of Montana, where the black rectangle indicates the location of the 14 study lakes, summers 2012 and 2013 (modified from Robbins, 2005) B A D C Figure 2.2: Measurements of body morphology for larval salamanders: A) snout-vent length (SVL), B) total length, C) tail length, and D) tail depth (modified from MacDonald, 2004)

45 Figure 2.3: Salamanders captured from a lake without trout (left) and a lake with trout (right), each matches the mean estimates of morphology for each lake type and are photographed at equal scales 35

46 Weight Total Length Tail Length (g) 1.0 (mm) 50 (mm) No Trout Trout No Trout Trout No Trout Trout 36 Figure 2.4: Body morphology of non-metamorphosing salamanders (means and 95% CIs) in no trout (n = 177 salamanders in seven lakes) and trout lakes (n = 1372 salamanders in seven lakes), northwestern Montana, summer 2012 and 2013

47 40 SVL No Trout 10 Tail Depth 0.95 Tail Index 35 Trout 9 (mm) 30 8 (mm) 7 (Tail/SVL) /27 8/1 8/6 8/11 8/16 8/21 8/26 8/31 7/27 8/1 8/6 8/11 8/16 8/21 8/26 8/31 7/27 8/1 8/6 8/11 8/16 8/21 8/26 8/31 37 Figure 2.5: Changes in body morphology over time of non-metamorphosing salamanders in no trout (n = 177 salamanders in seven lakes) and trout lakes (n = 1372 salamanders in seven lakes), northwestern Montana, summer 2012 and 2013 Date

48 Log SVL (mm) No Trout Trout Log Tail Length (mm) Log Tail Depth (mm) Log Weight (g) Figure 2.6: Allometric relationships between weight and SVL, tail length, and tail depth for non-metamorphosing salamanders captured in no trout (n = 177 salamanders in seven lakes) and trout lakes (n = 1372 salamanders in seven lakes), northwestern Montana, summer 2012 and

49 39 Probability of Metamorphosis No Trout Trout Day Figure 2.7: The probability of metamorphosing salamanders in trout (n = 52 of 1372 salamanders in seven lakes) and no trout lakes (n = 15 of 177 salamanders in seven lakes). Estimates are based on the number of metamorphosing individuals relative to the number of individuals captured over time, northwestern Montana, summer 2012 and 2013

50 40 Weight SVL Total Length* (g) (mm) 36 (mm) Tail Length* Tail Depth Tail Index (mm) 30 (mm) Tail Length/SVL No Trout Trout No Trout Trout No Trout Trout Figure 2.8: Comparisons of body morphology between metamorphosing salamanders (means and 95% CIs) captured in trout lakes (n = 16 salamanders in seven lakes) and no trout (n = 52 salamanders in seven lakes), northwestern Montana, summer 2012 and 2013 (* denotes significant difference

51 41 Literature Cited Abrams, P.A. and L. Rowe The effects of predation on the age and size of maturity of prey. Evolution 50: Adams, M.J., K.O. Richter, and W.P. Leonard Surveying and monitoring amphibians using aquatic funnel traps. Northwest Fauna 4: Altwegg, R. and H.U. Reyer Patterns of natural selection on size at metamorphosis in water frogs. Evolution 57: Anderson, J.D. and G.K. Williamson Nocturnal stratification in larvae of the mole salamander, Ambystoma talpoideum. Herpetologica 30: Arntzen, J Ecological observations on Chioglossa lusitanica caudate salamandridae. Amphibia-Reptilia 3-4: Arntzen, J.W Allometry and autotomy of the tail in the golden-striped salamander, Chioglossa lusitanica. Amphibia-Reptilia 15: Babbitt, K.J. and F. Jordon Predation on Bufo terrestris tadpoles, effects of cover and predator identity. Copeia 2: Bahls, P.F The status of fish populations and management of high mountain lakes in the western United States. Northwest Science 66: Barnett, H.K. and J.S. Richardson Predation risk and competition effects on the life-history characteristics of larval Oregon spotted frog and larval red-legged frog. Oecologia 132: Beck, C.W. and J.D. Congdon Effects of individual variation in age and size at metamorphosis on growth and survivorship of southern toad (Bufo terrestris) metamorphs. Canadian Journal of Zoology 77: Beckerman, A.P., M. Uriarte, and O.J. Schmitz Experimental evidence for a behavior-mediated trophic cascade in a terrestrial food chain. Proceedings of the National Academy of Sciences 94: Benard, M.F Predator-induced phenotypic plasticity in organisms with complex life histories. Annual Review of Ecology, Evolution, and Systematics 35: Berven, K.A The genetic basis of altitudinal variation in the wood frog Rana sylvatica. I. An experimental analysis of life history traits. Evolution 36:

52 42 Berven, K.A Factors affecting variation in reproductive traits within a population of wood frogs (Rana sylvatica). Copeia 1988: Bolnick, D.I. and E.I. Preisser Resource competition modifies the strength of traitmediated predator-prey interactions: A meta-analysis. Ecology 86: Boone, M.D., R.D. Semlitsch, E.E. Little, and M.C. Doyle Multiple stressors in amphibian communities: effects of chemical contamination, bullfrogs, and fish. Ecological Applications 17: Brönmark, C. and J.G. Miner Predator-induced phenotypical change in body morphology in crucian carp. Science 258: Brönmark, C. and L.A. Hansson The Biology of Lakes and Ponds. Oxford University Press, Oxford, England. Bull, E.L. and D.B. Marx Influence of fish and habitat on amphibian communities in high elevation lakes in northeastern Oregon. Northwest Science 76: Chivers, D.P., J.M. Kiesecker, A. Marco, E.L. Wildy, and A.R. Blaustein Shifts in life history as a response to predation in western toads (Bufo boreas). Journal of Chemical Ecology 25: Creel, S., J. Winnie Jr., B. Maxwell, K. Hamlin, and M. Creel Elk alter habitat selection as an antipredator response to wolves. Ecology 86: Daly, J.W The chemistry of poisons in amphibian skin. Proceedings of the National Academy of Sciences 92:9-13. Davenport, J.M., P.A. Seiwert, L.A. Fishback, and W.B. Cash The effects of two fish predators on wood frog (Lithobates sylvaticus) tadpoles in a subarctic wetland: Hudson Bay Lowlands, Canada. Canadian Journal of Zoology 91: Delgado-Acevedo, J. and C. Restrepo The contribution of habitat loss to changes in body size, allometry, and bilateral asymmetry in two Eleutherodactylus frogs from Puerto Rico. Conservation Biology 22: Dodd, M.H.I. and J.M. Dodd Physiology of Amphibia. Academic Press, United States. Duellman, W.E. and L. Trueb Biology of Amphibians. McGraw-Hill, United States.

53 Dunham, J.B., D.S. Pilliod, and M.K. Young Assessing the consequences of nonnative trout in headwater ecosystems in western North America. Fisheries 29: Emery, A.L., A.H. Berst, and K. Kodaira Under-ice observations of wintering sites of leopard frogs. Copeia 1: Figiel, C.R., Jr. and R.D. Semlitsch Population variation in survival and metamorphosis of larval salamanders (Ambystoma maculatum) in the presence and absence of fish predation. Copeia 1990: Figiel, C.R., Jr. and R.D. Semlitsch Effects of nonlethal injury and habitat complexity on predation in tadpole populations. Canadian Journal of Zoology 69: Finlay, J.C. and V.T. Vredenburg Introduced trout sever trophic connections in watersheds: consequences for a declining amphibian. Ecology 88: Funk, W.C. and W.W. Dunlap Colonization of high-elevation lakes by long-toed salamanders (Ambystoma macrodactylum) after the extinction of introduced trout populations. Canadian Journal of Zoology 77: Gervasi, S.S. and J. Foufopoulos Costs of plasticity: responses to desiccation decrease post-metamorphic immune function in a pond-breeding amphibian. Functional Ecology 22: Gilliam, J.F Foraging under mortality risk in size-structured populations. Ph.D. Dissertation, Michigan State University, East Lansing, MI. Gilliam, J.F. and D.F. Fraser Habitat selection when foraging under predation hazard: a model and a test with stream-dwelling minnows. Ecology 68: Grant, E.H.C Visual implant elastomer mark retention through metamorphosis in amphibians. Journal of Wildlife Management 75: Heyer, W.R., R.W. McDiarmid, and D.L. Weigmann Tadpoles, predation and pond habitats in the tropics. Biotropica 7: Hirner, J.L.M. and S.P. Cox Effects of rainbow trout (Oncorhynchus mykiss) on amphibians in productive recreational fishing lakes of British Columbia. Canadian Journal of Fish and Aquatic Sciences 64:

54 44 Huang, C. and A. Sih Experimental studies on direct and indirect interactions in a three tropic-level stream system. Oecologia 85: Joseph, M.B., J. Piovia-Scott, S.P. Lawler, and K.L. Pope Indirect effects of introduced trout on cascades frogs (Rana cascadae) via shared aquatic prey. Freshwater Biology 56: Kaplan, R.H. and S.N. Salthe The allometry of reproduction: an empirical view in salamanders. The American Naturalist 113: Kats, L.B., J.W. Petranka, and A. Sih Antipredator defenses and the persistence of amphibian larvae with fishes. Ecology 69: Kiesecker, J.M. and A.R. Blaustein Population differences in responses of redlegged frogs (Rana aurora) to introduced bullfrogs. Ecology 78: Kiesecker, J.M. and A.R. Blaustein Effects of introduced bullfrogs and smallmouth bass on microhabitat use, growth, and survival of native red-legged frogs (Rana aurora). Conservation Biology 12: Lardner, B Morphological and life history responses to predators in larvae of seven anurans. Oikos 88: Lawler, S.P., D. Dritz, T. Strange, and M. Holyoak Effects of introduced mosquitofish and bullfrogs on the threatened California red-legged frog. Conservation Biology 13: Lima, S.L Nonlethal effects in the ecology of predator-prey interactions. Bioscience 48: Ludwig, D. and L. Rowe Life-history strategies for energy gain and predator avoidance under time constraints. American Naturalist 135: Macan, T.T The influence of predation on the composition of fresh-water animal communities. Biological Review 52: Maiorana, V.C Tail autonomy, functional conflicts and their resolution by a salamander. Nature 265: McCauley, S.J., L. Rowe, and M.-J. Fortin The deadly effects of nonlethal predators. Ecology 92:

55 McDonald, S.O The amphibians and reptiles of Alaska: a field handbook. Available at: _sorted_for_printing.PDF 45 McNamara, J.M. and A.I. Houston The effect of a change in foraging options on intake rate and predation rate. American Naturalist 144: McPeek, M.A. and B.L. Peckarsky Life histories and the strengths of species interactions: combining mortality, growth, and fecundity effects. Ecology 79: McPeek, M.A The growth/predation risk trade-off: so what is the mechanism? American Naturalist 163:E88-E111. Mittelbach, G.G Community Ecology. Sinauer Associates, United States Montana Fish, Wildlife & Parks Fish Stocking Report. Available at: Montana Fish, Wildlife & Parks Fish Stocking Report. Available at: Montana Fish, Wildlife & Parks Fish Stocking Report. Available at: Montana Fish, Wildlife & Parks Fish Stocking Report. Available at: Montana Fish, Wildlife & Parks Fish Stocking Report. Available at: Newman, R.A Effects of changing density and food level on metamorphosis of a desert amphibian, Scaphiopus couchii. Ecology 75: Nyström, P., O. Svensson, B. Lardner, C. Brönmark, and W. Granéli The influence of multiple introduced predators on a littoral pond community. Ecology 82: Orizaola, G. and F. Brana Plasticity in newt metamorphosis: the effect of predation at embryonic and larval stages. Freshwater Biology 50: Paoletti, D.J., D.H. Olsen, and A.R. Blaustein Responses of foothill yellow-legged frog (Rana boylii) larvae to an introduced predator. Copeia 2011:

56 46 Pearl, C.A., M.J. Adams, G.S. Schuytema, and A.V. Nebeker Behavioral responses of anuran larvae to chemical cues of native and introduced predators in the pacific northwestern United States. Journal of Herpetology 37: Pearson, K.J. and C.P. Goater Distribution of long-toed salamanders and introduced trout in high- and low-elevation wetlands in southwestern Alberta, Canada. Ecoscience 15: Pearson, K.J. and C.P. Goater Effects of predaceous and nonpredaceous introduced fish on the survival, growth, and antipredation behaviours of long-toed salamanders. Canadian Journal of Zoology 87: Peckarsky, B.L., B.W. Taylor, A.R. McIntosh, M.A. McPeek, and D.A. Lytle Variation in mayfly size at metamorphosis as a developmental response to risk of predation. Ecology 82: Peckarsky, B.L., A.R. McIntosh, B.W. Taylor, and J. Dahl Predator chemical induce changes in mayfly life history traits: a whole stream manipulation. Ecology 83: Petranka, J.W., L.B. Kats, and A. Sih Predator-prey interactions among fish and larval amphibians: use of chemical cues to detect predatory fish. Animal Behaviour 35: Petranka, J.W Density-dependent growth and survival of larval Ambystoma: evidence from whole-pond manipulations. Ecology 70: Pilliod, D.S. and C.R. Peterson Local and landscape effects of introduced trout on amphibians in historically fishless watersheds. Ecosystems 4: Pilliod, D.S., B.R. Hossack, P.F. Bahls, E.L. Bull, and P.S. Corn Non-native salmonids affect amphibian occupancy at multiple spatial scales. Diversity and Distributions 16: Pilliod, D.S., R.S. Arkle, and B.A. Maxell Persistence and extirpation in invaded landscapes: patch characteristics and connectivity determine effects of non-native predatory fish on native salamanders. Biological Invasions 15: Preisser, E.L., D.L. Bolnick, and M.F. Benard Scared to death? The effects of intimidation and consumption in predator-prey interactions. Ecology 86: R Development Core Team R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN , Available at:

57 47 Robbins, C Flyfisher s Guide to Montana. Wilderness Adventures Press, Canada. Rowe, L. and D. Ludwig Size and timing of metamorphosis in complex life cycles time constraints and variation. Ecology 72: Salo, P., E. Korpimäki, P.B. Banks, M. Nordström, and C.R. Dickman Alien predators are more dangerous than native predators to prey populations. Proceedings of the Royal Society B. 274: Schmitz, O.J Direct and indirect effects of predation and predation risk on oldfield interaction webs. The American Naturalist 151: Scott, D.E Effects of larval density in Ambystoma opacum: an experiment largescale field enclosures. Ecology 71: Semlitsch, R.D. and J.P. Caldwell Effects of density of growth, metamorphosis, and survivorship in tadpoles in Scaphiopus holbrooki. Ecology 63: Semlitsch, R.D Interactions between fish and salamander larvae. Costs of predator avoidance or competition? Oecologia 72: Semlitsch, R.D., D.E. Scott, and J.H.K. Pechmann Time and size at metamorphosis related to adult fitness in Ambystoma talpoideum. Ecology 69: Sih, A Prey refuges and predator-prey stability. Theoretical Population Biology 31:1-12. Sih, A., J.W. Petranka, and L.B. Kats The dynamics of prey refuge use: a model and tests with sunfish and salamander larvae. The American Naturalist 132: Skelly, D.K. and E.E. Werner Behavioral and life-historical responses of larval American toads to an odonate predator. Ecology 71: Skelly, D.K Field evidence for a cost of behavioral avoidance response in a larval amphibian. Ecology 73: Skelly, D.K A behavioral trade-off and its consequences for the distribution of Pseudacris treefrog larvae. Ecology 76: Smith, D.C Adult recruitment in chorus frogs: effects of size and date at metamorphosis. Ecology 68:

58 48 Stauffer, H. and R.D. Semlitsch Effects of visual, chemical and tactile cues of fish on the behavioural responses of tadpoles. Animal Behavior 46: Stenhouse, S.L Interdemic variation in predation on salamander larvae. Ecology 66: Stuart, S.N., J.S. Chanson, N.A. Cox, B.E. Young, A.S.L. Rodrigues, D.L. Fischman, and R.W. Waller Status and trends of amphibian declines and extinctions worldwide. Science 306: Taylor, J Orientation and flight behavior of a neotenic salamander (Ambystoma gracile) in Oregon. American Midland Naturalist 109: Taylor, B., D. Skelly, L.K. Demarchis, M.D. Slade, D. Galusha, and P.M. Rabinowitz Proximity to pollution sources and risk of amphibian limb malformation. Environmental Health Perspectives 113: Tyler, T.J., W.J. Liss, R.L. Hoffman, and L.M. Ganio. 1998a. Experimental analysis of trout effects on survival, growth, and habitat use of two species of Ambystomatid salamanders. Journal of Herpetology 32: Tyler, T., W.J. Liss, L.M. Ganio, G.L. Larson, R. Hoffman, E. Deimling, and G. Lomnicky. 1998b. Interaction between introduced trout and larval salamanders (Ambystoma macrodactylum) in high-elevation lakes. Conservation Biology 12: Van Buskirk, J. and D.C. Smith Density-dependent population regulation in a salamander. Ecology 72: Van Buskirk, J. and B.R. Schmidt Predator-induced phenotypic plasticity in larval newts: trade-offs, selection, and variation in nature. Ecology 81: Warne, R.W., E.J. Crespi, and J.L. Brunner Escape from the pond: stress and developmental responses to ranavirus infection in wood frog tadpoles. Functional Ecology 25: Welsh, H.H., Jr., K.L. Pope, and D. Boiano Sub-alpine amphibian distributions related to species palatability to non-native salmonids in the Klamath mountains of northern California. Diversity and Distributions 12: Werner, E.E., J.F. Gilliam, D.J. Hall, and G.G. Mittelbach An experimental test of the effects of predation risk on habitat use in fish. Ecology 64:

59 49 Werner, E.E. and S.D. Peacor A review of trait-mediated indirect interactions in ecological communities. Ecology 84: Wilbur, H.M. and J.P. Collins Ecological aspects of amphibian metamorphosis. Science 182: Yurewicz, K.L A growth/mortality trade-off in larval salamanders and the coexistence of intraguild predators and prey. Oecologia 138: Zuur, A., E.N. Ieno, N. Walker, A.A. Saveliev, and G.M. Smith Mixed effects models and extensions in ecology with R. Springer, United States.

60 50 CHAPTER THREE ADDING VEGETATION STRUCTURE TO REDUCE NONCONSUMPTIVE EFFECTS OF INTRODUCED TROUT: A NOVEL METHOD FOR AMPHIBIAN CONSERVATION? Contributions of Authors and Co-Authors Manuscript in Chapter 3 Author: Erin K. Kenison Contributions: Conceived the study, collected and analyzed data, and authored the manuscript. Co-author: Andrea R. Litt Contributions: Conceived the study, assisted with study design, and obtained funding. Discussed results, aided in interpretation of analyses, and edited manuscript. Co-author: David S. Pilliod Contributions: Helped conceive the study and assisted with study design. Discussed interpretation and implications of results and reviewed manuscript. Co-author: Thomas E. McMahon Contributions: Helped conceive the study and assisted with study design. Discussed interpretation and implications of results and reviewed manuscript.

61 51 Manuscript Information Page Erin K. Kenison, Andrea R. Litt, David S. Pilliod, and Thomas E. McMahon Journal of Herpetology Status of Manuscript: x_ Prepared for submission to a peer-reviewed journal Officially submitted to a peer-reviewed journal Accepted by a peer-reviewed journal Published in a peer-reviewed journal Published by The Society for the Study of Amphibians and Reptiles

62 52 CHAPTER THREE ADDING VEGETATION STRUCTURE TO REDUCE NONCONSUMPTIVE EFFECTS OF INTRODUCED TROUT: A NOVEL METHOD FOR AMPHIBIAN CONSERVATION? Abstract Predators regulate prey populations, but introduced predators can have a greater impact on native prey. Trout introduced to historically fishless lakes have led to reduced abundance and even local extinctions of amphibian populations. Although uncommon, trout and larval amphibians can coexist in lakes with complex vegetation. Larvae can detect trout cues and rely on increased refuge use to reduce encounter rates with fish predators. However, these behavioral responses can result in decreases in energy uptake and body size, as well as survival and reproduction (nonconsumptive effects). We conducted a field experiment to investigate whether adding vegetation structure can reduce the nonconsumptive effects of trout on long-toed salamander larvae. We constructed field enclosures in lakes with and without trout and quantified changes in salamander growth, probability of metamorphosis, and size at metamorphosis among four treatment types: presence/absence of trout cues and presence/absence of structure. Salamanders appeared to detect trout cues because they grew more slowly in lakes with trout, even though trout had no ability to consume larvae. Added vegetation structure did not influence growth rates, but did increase the probability of salamanders that reached metamorphosis before the end of the experiment. Salamanders did not differ in size at

63 53 metamorphosis across all four treatments, but slower growth and reduced probability of metamorphosis may negatively impact persistence of salamander populations. Future research should investigate whether adding vegetation structure to the whole lake can facilitate coexistence and provide a feasible alternative to trout removal. Introduction Predator-prey interactions have the capacity to shape population dynamics, species distributions, community composition, and ultimately, regulate ecosystem function (Menge and Sutherland, 1976; Kerfoot and Sih, 1987). Depending on the efficiency of a predator at capturing and consuming prey, predator and prey populations can oscillate in an unstable manner over time (Rosenzweig and MacArthur, 1963). Instability in predator-prey interactions can lead to population booms or increased risk of extinction for prey, predators, or both (Stiling, 2002). Nonnative predators can have a much greater impact on prey compared to native predators, exacerbating this instability (Salo et al., 2007). The introduction of nonnative species has been associated with global amphibian declines and is considered the greatest threat to freshwater ecosystems (Collins and Storfer, 2003; Kats and Ferrer, 2003). Native prey species are especially vulnerable because they often are naïve to nonnative predators, may not be able to detect introduced predators or do not possess effective adaptations or learned behavior to avoid predation (Kiesecker and Blaustein, 1997; Kats and Ferrer, 2003). Pacific treefrogs (Pseudacris regilla), for example, respond to native fish predators by increasing refuge use, but are

64 54 unable to detect chemical cues of nonnative fish and don t increase time spent under cover (Pearl et al., 2003). Red-legged frog (Rana aurora) tadpoles that are naïve to bullfrogs (Rana catesbeiana), do not alter behavior when exposed to bullfrogs and are more likely to be depredated, whereas syntopic tadpoles decrease activity and increase refuge use in response to bullfrog larvae and adults (Kiesecker and Blaustein, 1997). Neurotoxins in the skin of adult newts (Taricha torosa) are effective defenses against native predators, but are insufficient to protect against predation by introduced crayfish (Procambarus clarkia, Gamradt et al., 1997). Without appropriate responses to introduced predators, populations of amphibians are more likely at risk of decline (Pearl et al., 2003). Throughout much of the western United States, trout have been introduced to historically fishless lakes for recreational purposes (Bahls, 1992). These introductions have resulted in decreases in abundance and even extirpations of populations of amphibians in lakes where trout exist (Funk and Dunlap, 1999; Pilliod and Peterson, 2001; Bull and Marx, 2002). Trout prey on eggs, larvae/tadpoles, juveniles, and occasionally adult amphibians (Matthews et al., 2001; Welsh et al., 2006; Wells, 2007). Most amphibian larvae are capable of detecting fish predator cues, but must rely on changes in behavior to reduce predation risk when they lack chemical or morphological defenses (Petranka et al., 1987; Tyler et al., 1998a; Pearson, 2004). When exposed to introduced trout, larval amphibians increase refuge use and decrease time spent in open water to reduce encounter rates (Petranka et al., 1987; Kats et al., 1988; Tyler et al., 1998a). Although these changes in behavior may increase amphibian survival and reduce

65 55 predation, fish predators still can have nonconsumptive effects on larvae (Tyler et al., 1998a; Hirner and Cox, 2007; Chapter 2). As individuals increase the amount of time spent in refugia, foraging activity decreases, with concomitant decreases in energy uptake and body size (Semlitsch, 1987; Figiel and Semlitsch, 1990; Skelly and Werner, 1990; Tyler et al., 1998a; Chapter 2). Smaller body sizes during the larval period and at metamorphosis are associated with reduced survival and reproduction and increased susceptibility to terrestrial threats such as predation, desiccation, and disease (Petranka et al., 1987; Semlitsch, 1987; Figiel and Semlitsch, 1990; Kiesecker and Blaustein, 1998; Tyler et al., 1998a; Gervasi and Foufopoulos, 2008; Warne et al., 2011). The long-toed salamander (Ambystoma macrodactylum) is the most widely distributed amphibian west of the Continental Divide, extending from central California as far north as southeast Alaska, and is the most common salamander in the state of Montana (Werner et al., 2004). However, this species is especially vulnerable to introduced fish due to their palatability, lack of chemical defenses, 1-3 year larval period, and reliance on deep lakes for overwintering (Howard and Wallace, 1985; Pilliod and Peterson, 2000; Werner et al., 2004). Salamander larvae respond to predator cues by increasing use of refugia to avoid predation, yet in lakes with trout, abundance of salamander egg masses is lower, larvae are >65% less abundant, and some populations have been extirpated (Tyler et al., 1998b; Funk and Dunlap, 1999; Pilliod and Peterson, 2001; Bull and Marx, 2002; Hirner and Cox, 2007; Pearson and Goater, 2008).

66 56 Furthermore, salamander occupancy is negatively correlated with fish presence and coexistence between trout and salamanders occurs rarely (Pilliod et al., 2013). Removing fish from historically fishless lakes may be the most effective way to reverse effects on long-toed salamander populations. When fish are removed from lakes, amphibians recolonize sites and population sizes increase rapidly, reducing the risk of local extinction (Funk and Dunlap, 1999; Knapp et al., 2001; Hoffman et al., 2004; Vredenburg, 2004). However, alternative conservation strategies are needed, given that removing trout often is unrealistic due to high costs and effort, in addition to being unpopular with the public. Dense or structurally-complex vegetation or cover objects provide protection from predators and can facilitate coexistence between predators and prey (Sih, 1987; Allouche, 2002). Complex structure reduces the maneuverability, visual range, and effectiveness of fish predators and has been correlated with increased prey survival (Werner et al., 1983; McNair, 1986; Sih, 1987; Kats et al., 1988; Figiel and Semlitsch, 1991; Diehl, 1992; Babbitt and Jordan, 1996; Babbitt and Tanner, 1997). Despite the rarity of coexistence, long-toed salamanders and trout have been found to co-occur in lakes where habitat characteristics such as emergent vegetation and physical barriers from trout are present (Tyler et al., 1998a; Pearson and Goater, 2008; Pilliod et al., 2010; Pilliod et al., 2013). Therefore, adding complex vegetation structure, such as woody debris, to lakes with trout could provide an alternative strategy for conservation of salamanders. Previous amphibian conservation projects have been directed at maintaining buffer zones surrounding breeding sites, creating connectivity between wetlands, or

67 57 constructing and restoring water bodies (Semlitsch, 2002; Shulse et al., 2012), yet, adding complex structure may create refugia that can reduce exposure to predators, provide protected areas to forage, and dampen nonconsumptive effects of trout, potentially increasing fitness of terrestrial adults (Bull and Marx, 2002; Relyea, 2007; Shulse et al., 2012). We sought to understand how adding complex vegetation structure affects growth, body size, and timing of metamorphosis of larval long-toed salamanders in lakes with and without trout. We aimed to determine if this strategy could reduce the nonconsumptive effects of introduced trout and provide an alternative conservation strategy to fish removal. We hypothesized that salamanders would grow faster and be larger at metamorphosis in lakes without trout, but that salamanders in lakes with trout and added structure would be comparable in size. Additionally, we hypothesized that the timing of salamander metamorphosis would be earlier in lakes without trout and that salamanders in enclosures with added structure would metamorphosis faster than those without. If salamanders can grow and metamorphose early and at larger sizes, adults may have higher rates of survival and reproduction, ultimately increasing size of salamander populations. We know of no other studies that have experimentally manipulated cover to assess if added structural complexity could enhance amphibian coexistence with trout predators. We seek to test the effectiveness of this management strategy and its potential use in conservation efforts for other amphibian populations negatively affected by introduced predators.

68 58 Methods Study Sites: We conducted a field experiment during the summer of 2013 in five permanent lakes, two lakes with trout and three without trout. Lakes were located in northwestern Montana, in the Flathead National Forest between the Swan and Mission Mountain ranges; mainly between Condon and the town of Seeley Lake, with one lake located closer to Missoula, Montana (Table 3.1). We selected all lakes during previous visual encounter and trapping surveys in 2012 that confirmed the presence of salamanders and the presence or absence of trout (Chapter 2). Field Experiment: We constructed experimental field enclosures with PVC pipe and fiberglass window screening (2 x 1 x 1 m, LWH) (Sredl and Collins, 1991; Kiesecker and Blaustein, 1998). Experimental treatments were based on a 2 x 2 factorial arrangement: 1) presence or absence of trout cues (i.e., enclosures in lakes with trout compared to those without trout) and 2) presence or absence of added vegetation structure. This design exposed salamanders to the chemical or visual signals of predators, but protected them from direct predation. We randomly assigned treatments to enclosures. We constructed complex vegetation structures by binding roughly 20 dry, dead sticks with 20 living, leafy branches (i.e., aspen or other littoral vegetation) with twine in an inverted funnel shape (Schneider and Winemiller, 2008, Figure 3.1). Vegetation structures were close to 1 m in circumference at the base and cm tall. We placed

69 59 two vegetation structures in enclosures assigned added structure treatments, one structure at each end of the enclosure, and secured them in place with a large rock. We placed enclosures along the north-facing edge of each lake in the littoral zone. We positioned enclosures so that depth, distance between enclosures, and distance from shore was equal for all enclosures (Sredl and Collins, 1991). We also added ~10 L of lake substrate to the bottom of each enclosure to incorporate natural sediments. Lakes with trout had 4 replicates of each of the 2 treatments (trout cues, with and without added vegetation structure), for a total of 8 enclosures per lake. Lakes without trout had 2 replicates of each of the 2 treatments (absence of trout cues, with and without added vegetation structure), for a total of 4 enclosures per lake. We used minnow traps to capture salamanders to add to experimental enclosures during June and July We used minnow traps, as they are especially useful in trapping small, cryptic larvae and are associated with little mortality or injury (Adams et al., 1997). We anesthetized salamander larvae with MS-222, measured weight, snout-vent length (SVL), total length, tail length, and tail depth with an electronic scale and calipers, and marked each individual uniquely with visual implant elastomer in the middle of their tail. We randomly assigned 20 salamanders to each enclosure; salamanders were comparable in size, >0.2 g, and we marked and measured individuals before adding them to enclosures. We selected this density of salamanders as it was comparable to previous studies that did not observe cannibalism and where adequate amounts of food persisted throughout study periods (e.g., Clark, 1986; Semlitsch, 1987; Tyler et al., 1998a; Pearson and Goater, 2009). Although we expected zooplankton and other small crustaceans to be

70 60 able to pass through enclosure screen, we added ~10 L of lake water to enclosures at each visit to provide additional food resources (Stenhouse, 1985; Tarr and Babbitt, 2002). After adding salamanders to enclosures, we visited enclosures four additional times during July and August At each visit, we captured individual larvae with hand nets, recorded measurements of body morphology (i.e., weight, SVL, total length, tail length, and tail depth), noted whether salamanders were metamorphosing, which we defined as any evidence of gill absorption (Dodd and Dodd, 1976; Arntzen, 1981; Duellman and Trueb, 1986), and released them back into assigned enclosures. We were unable to accurately measure survival, as we couldn t distinguish with certainty whether salamanders missing from enclosures escaped or died from predation by invertebrates, conspecific cannibalism, or natural causes. We terminated the experiment at the end of August, released all remaining individuals, and removed enclosures. We anesthetized, marked, and handled all captured individuals in accordance with Montana State University Institutional Animal Care and Use Committee protocol Statistical Analyses: We assessed the effects of trout cues on salamanders in enclosures with and without added vegetation structure by comparing changes in morphological measurements over time among all four treatments: 1) presence of trout cues (i.e., lakes with trout) and added structure, 2) presence of trout cues and no added structure, 3) absence of trout cues (i.e., lakes without trout) and added structure, and 4) absence of trout cues and no added structure. We also examined differences in the probability of

71 61 metamorphosis, timing of metamorphosis, and size of salamanders at metamorphosis among all four treatments. We analyzed all measurements of body morphology: weight, SVL, total length, tail length, and tail depth as separate response variables for non-metamorphosing and metamorphosing salamanders. We used generalized linear mixed models, random slopes and intercepts, and a three-tiered, nested data structure multiple measurements of salamanders within enclosures within study lakes to account for repeated measurements and variation among lakes (Zuur et al., 2009). We included our experimental treatments presence/absence of trout (trout) and presence/absence of vegetation structure (structure) as simple effects in analytical models. We also included time in models as number of days since the first sampling visit (day), as we hypothesized that changes in morphological measurements over time might differ based on vegetation structure and trout condition. We considered all possible 2-way interactions between treatment factors and time (trout * structure, trout * day, structure * day) in models of larval morphology and probability of metamorphosis. We assessed evidence of effects of trout and structure on size at metamorphosis by examining simple (trout, structure) and interactive effects (trout * structure). We removed interactions that did not explain sufficient variation (α > 0.1) from models for inference, but kept all simple explanatory variables in the model. We calculated relative difference over time by dividing the average rate of growth in enclosures without trout cues over the average rate of growth in enclosures with trout cues. We report estimates of changes in morphological measurements over time and average size at metamorphosis, and include 95% confidence intervals for all estimates.!

72 62 Results At the beginning of the experiment, salamanders were similar in size among treatments (see differences in intercepts of trout and structure, Table 3.2). All body morphology measurements of non-metamorphosing salamanders increased over time (trout * day, Table 3.2). Changes in all morphological measurements were greater in lakes without trout compared to lakes with trout and did not depend on structure (trout * structure, Figure 3.2) On average, growth rates for salamander in enclosures without trout increased faster by 2.9 times for weight, 2.2 times for SVL, 2.4 times for total length, 2.3 times for tail length, and 6.7 times for tail depth (Figure 3.2, Table 3.3). Adding vegetation structure was not influential for most of the body morphology measurements, but did slow the growth of tail depth in enclosures with and without trout cues (structure * day, Figure 3.3, Table 3.2). Weight, SVL, total length, or tail depth of metamorphosing salamanders were similar with and without trout cues and with and without vegetation structure (Table 3.4). However, salamanders in enclosures with added structure had tail lengths that were 4.6% shorter (95% CI = %) than salamanders without added structure, regardless of the presence or absence of trout cues (Table 3.4). Although salamanders were similar in size, the probability of metamorphosis over time did differ among treatments (Table 3.5). Adding structure to enclosures with trout cues increased the probability of metamorphosis to be similar to enclosures without trout cues; probability of metamorphosis was lowest in enclosures with trout cues that did not have added structure (Figure 3.3).

73 63 Discussion Structural complexity is important in mediating species interactions and facilitating the coexistence of predators and prey (Murdoch and Oaten, 1975; Sih, 1987; Diehl, 1992). Coexistence between trout and salamanders is uncommon, often only occurring in lakes with emergent vegetation and shallow littoral zones, areas that provide adequate separation and refuge from fish predators (Tyler et al., 1998b; Pearson and Goater, 2008; Pilliod et al., 2010, 2013). The presence of refugia can stabilize predatorprey interactions, as increased complex vegetation reduces the frequency and success of predator attacks on prey. However, prey still can incur nonconsumptive costs with the use of refugia (Rosenzweig and MacArthur, 1963; Diehl, 1988). We provide evidence that trout chemical cues induce nonconsumptive effects that alter long-toed salamander growth and these effects are more influential on the size of salamanders than the addition of complex vegetation structure. Trout clearly influenced salamander growth in our study, as larvae grew more slowly in lakes with trout compared to lakes without trout (also see Chapter 2). Even without the physical risk of predation, chemical, visual, or both types of trout cues are sufficient for salamanders to detect and respond with predator avoidance behaviors, such as reduced foraging (Stauffer and Semlitsch, 1993). Tadpoles subjected to the presence of a caged predator or fish chemical cues demonstrate as much as a 41% reduction in foraging or a 68% reduction in time spent outside of refugia (Petranka et al., 1987; Skelly and Werner, 1990). Persisting in environments with fish that are perceived as high-risk not only reduces salamanders ability to acquire resources, but also increases stress

74 64 levels, which can inhibit food intake, suppress appetite, and ultimately, reduce size and growth over the larval period (Crespi and Denver, 2005). Although vegetation complexity is important for, and increases the likelihood of coexistence (Sih, 1987), adding vegetation structure did not mitigate the effects of trout cues on larval growth or size at metamorphosis in our experiment. However, adding structure to enclosures with trout cues did increase the probability of metamorphosis. Decreased activity in the presence of predators can reduce growth rates and lengthen larval periods (Nicieza, 2000; Babbitt, 2001; Johansson et al., 2001; Altwegg, 2002). Organisms often metamorphose when greater resources can be acquired and increased fitness can be achieved as an adult rather than larvae, but a minimum size must be reached before metamorphosis can be initiated (Wilbur and Collins, 1973; Gilliam, 1982; Benard, 2004). Early metamorphosis can occur under heighted risk of predation, for example western toads shortened their time to metamorphosis by ~10 days when exposed to predatory cues of backswimmers (Notonecta spp., Chivers et al., 1999). Alternatively, in our study, reduced probability of metamorphosis in enclosures with trout cues but without added structure suggests that salamanders may detect risks associated with trout, become more vigilant without adequate sources of refuge but are less likely to leave the aquatic system as early as salamanders without trout cues or with added structure (Nicieza, 2000; Babbitt, 2001; Altwegg, 2002). Salamanders did not differ in size at metamorphosis, despite differences in the probability of metamorphosis. Southern leopard frogs, wood frogs, and water frogs extend their larval period when exposed to insect predators, but metamorphose at equal

75 65 sizes (Relyea, 2001; Babbitt, 2001; Van Buskirk and Saxer, 2001). Similarly, long-toed salamanders have longer larval periods and reduced growth rates when in the presence of cannibalistic conspecifics, yet are similar in size at metamorphosis to larvae not exposed to predators (Wildy et al., 1999). Salamanders that remain in their larval period longer in lakes with trout may increase their exposure to predators, reducing the number of individuals that leave the pond and successfully enter into the terrestrial system as adults. The combination of nonconsumptive and consumptive effects of trout ultimately may influence populations if reductions in adult survival and reproduction limit recruitment (Kiesecker and Blaustein, 1998; Nicieza, 2000). Future research should focus on whether adding complex vegetation structure at a whole-lake level can be effective at reducing the nonconsumptive effects of trout. By adding vegetation structures to lakes, instead of the restricted space of an enclosure, and allowing structures to remain over sufficient time for invertebrate prey to establish, we might be able to create highly productive, food-rich microhabitats for larval amphibians (Diehl, 1992; Hartel, 2004). Habitat features such as rocks, woody material, and vegetation are effective protection from fish predators that also are associated with increased density and species richness of invertebrates (Stenhouse, 1985; Babbitt and Jordan, 1996; Tyler et al., 1998b; Hartel et al., 2007). These areas of structural and vegetative complexity may provide food as well as safety from predators during larval development, which may mitigate nonconsumptive effects of trout even when predator avoidance behaviors are used. Alternative strategies to reduce the effects of introduced trout are essential,

76 66 especially as the number of amphibian species that are threatened and endangered continue to increase and populations of the even most common species decrease in size (Adams et al., 2013). Changes to amphibian populations can alter community composition and lead to disruptions of food webs in both the aquatic and terrestrial system (Kats and Ferrer, 2003). Introduced trout in the Sierra Nevada Mountains reduce the availability of aquatic insect prey emerging from lakes, resulting in decreases in the abundance of foraging frogs, and subsequent decreases in abundance of native garter snakes (Matthews et al., 2002; Kats and Ferrer, 2003; Finlay and Vredenburg, 2007). With additional research, adding vegetation complexity has the potential to efficiently improve persistence of amphibian species, facilitate coexistence between introduced predators and native prey, and preserve the function of terrestrial and aquatic ecosystems without requiring removal of fish.

77 67 Tables Table 3.1: Locations and characteristics of study lakes with (n = 2) and without trout (n = 3), northwestern Montana, summer 2013 Lake Latitude Longitude Trout Perimeter (m) Elevation (m) Elk No New No Pee No Rock Yes Spook Yes

78 Table 3.2: Factors affecting measurements of body morphology of salamanders, 20 per enclosure with repeated measures, in experimental enclosures without trout cues or added structure (n =6 enclosures), without trout cues, but with added structure (n = 6 enclosures), with trout cues, but without added structure (n = 8 enclosures), and with trout cues and added structure (n = 8 enclosures), and northwestern Montana, summer 2013 Weight SVL Total Length Explanatory df t P t P t P Trout Structure Day < < <0.001 Trout * Day < <0.001 Tail Length Tail Depth df t P t P Trout Structure Date < <0.001 Trout * Day <0.001 Structure * Day

79 Table 3.3: Estimated change in salamander body morphology over time (means and 95% CIs) in enclosures with/without trout cues (+/ ) and with/without added structure (+/ ), northwestern Montana, summer 2013 Response + Trout + Structure + Trout Structure Trout + Structure Trout Structure Weight (Δ g/10 days) SVL (Δ mm/10 days) Total Length (Δ mm/10 days) Tail Length (Δ mm/10 days) Tail Depth (Δ mm/10 days) 0.10 ( ) 1.51 ( ) 2.68 ( ) 1.33 ( ) 0.04 ( ) 0.12 ( ) 1.71 ( ) 2.87 ( ) 1.35 ( ) 0.14 ( ) 0.31 ( ) 3.43 ( ) 6.54 ( ) 3.11 ( ) 0.55 ( ) 0.33 ( ) 3.63 ( ) 6.74 ( ) 3.13 ( ) 0.65 ( ) 69

80 Table 3.4: Factors affecting body morphology of salamanders at metamorphosis in experimental enclosures without trout cues or structure (n = 18 salamanders in six enclosures), without trout cues, but with added structure (n = 24 salamanders in six enclosures), with trout cues, but without structure (n = 36 salamanders in eight enclosures), and with trout cues and added structure (n = 47 salamanders in eight enclosures), northwestern Montana, summer 2013 Weight SVL Total Length Explanatory df t P t P t P Trout Structure Tail Length Tail Depth df t P t P Trout Structure

81 71 Table 3.5: Factors affecting the probability of metamorphosis in experimental enclosures without trout cues or structure (n = 18 salamanders in six enclosures), without trout cues, but with added structure (n = 24 salamanders in six enclosures), with trout cues, but without structure (n = 36 salamanders in eight enclosures), and with trout cues and added structure (n = 47 salamanders in eight enclosures), northwestern Montana, summer 2013 Probability of Metamorphosis Explanatory df t P Trout Structure Day <0.001 Trout * Structure Trout * Day <0.001 Structure * Day Trout * Structure * Day

82 72 Figures Figure 3.1: Vegetation structure constructed for enclosures, made with dead and living plant materials and secured with twine